Table 1. Correlation in timing, weekly incidence and seasonal intensity of influenza epidemics, as measured by 3 surveillance systems: IMS-ILI, CDC-ILI, and CDC laboratory surveillance, by region from 2003–04 to 2009–10.
| Correlation outcome | Region 1 (Boston) | Region 2 (New York City) | Region 3 (Wash. DC) | Region 4 (Atlanta) | Region 5 (Chicago) | Region 6 (Dallas) | Region 7 (Kansas City) | Region 8 (Denver) | Region 9 (San Francis.) | Region 10 (Seattle) | Avg. (95% CI) |
| IMS-ILI v. CDC-ILI | |||||||||||
| Wkly inc. (lag)a | 0.94 (1) | 0.88 (0) | 0.90 (1) | 0.93 (1) | 0.97 (1) | 0.90 (1) | 0.94 (1) | 0.90 (1) | 0.90 (1) | 0.83 (0) | 0.91 (0.88; 0.93) |
| Peak week | 0.93 | 0.95 | 0.87 | 0.94 | 0.98 | 0.95 | 0.99 | 0.70 | 0.99 | 0.80 | 0.91 (0.85; 0.97) |
| Seas. intensity | 0.92 | 0.63 | 0.24 | 0.62 | 0.62 | 0.09 | 0.76 | 0.19 | 0.54 | 0.54 | 0.52 (0.33; 0.68) |
| IMS-ILI v. CDC viral surveillance | |||||||||||
| Wkly inc. (lag)b | 0.89 (1) | 0.87 (1) | 0.89 (1) | 0.93 (1) | 0.85 (0) | 0.90 (1) | 0.88 (0) | 0.93 (1) | 0.80 (1) | 0.94 (1) | 0.89 (0.86; 0.91) |
| Peak week | 0.93 | 0.97 | 0.97 | 0.92 | 0.94 | 0.99 | 0.99 | 0.98 | 0.99 | 0.97 | 0.97 (0.95; 0.99) |
| Seas. intensity | 0.06 | 0.15 | 0.05 | −0.02 | −0.27 | 0.66 | 0.23 | 0.34 | −0.18 | 0.10 | 0.11 (−0.05; 0.28) |
| CDC-ILI v. CDC viral surveillance | |||||||||||
| Wkly inc. (lag)c | 0.91 (0) | 0.83 (0) | 0.83 (0) | 0.92 (0) | 0.85 (0) | 0.86 (0) | 0.89 (0) | 0.90 (0) | 0.71 (1) | 0.82 (1) | 0.85 (0.81; 0.89) |
| Peak week | 0.98 | 0.86 | 0.94 | 0.96 | 0.96 | 0.95 | 0.97 | 0.76 | 0.99 | 0.79 | 0.92 (0.87; 0.97) |
| Seas. intensity | 0.35 | 0.54 | 0.62 | 0.58 | −0.02 | 0.33 | 0.75 | 0.81 | 0.38 | 0.52 | 0.49 (0.34; 0.63) |
Values indicate Pearson correlation coefficients; values in bold are significant. For ILI time series, seasonal intensity is based on excess incidence over baseline each season, estimated from Serfling seasonal regression. For CDC laboratory surveillance, seasonal intensity is based on the cumulative percent virus positive each season (sum of influenza virus positives/sum of respiratory specimens tested).
a
Lag maximizing the correlation between the two indicators indicated in parentheses. A positive lag indicates that IMS-ILI surveillance is ahead of CDC-ILI.
b
A positive lag indicates that IMS-ILI is ahead of CDC viral surveillance.
c
A positive lag indicates that CDC-ILI surveillance is ahead of CDC laboratory-confirmed viral activity.